David Ulevitch of Andreessen Horowitz joins Nick to discuss American Dynamism: The Future of U.S. Industrials, Backing Companies with Major Production Components, Manufacturing Sovereignty, and Why Space Dominance is Critical. In this episode we cover:
- Challenges in Venture Capital and Investment Philosophy
- Handling Startups and Market Pivots
- Navigating Dual-Use Startups
- Government Sales and Market Education
- Long-Term Revenue and Production Challenges
- American Dynamism Practice and Investment Thesis
- Supply Chain and Vertical Integration
- Policy Advocacy and Government Affairs
- Future of American Dynamism and Energy Investments
Guest Links:
The host of The Full Ratchet is Nick Moran of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.
We’re proud to partner with Ramp, the modern finance automation platform. Book a demo and get $150—no strings attached.
Want to keep up to date with The Full Ratchet? Follow us on social. You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.
0:17
David Ulevitch joins us today from New York City. He’s a General Partner at Andreessen Horowitz, where he leads the American Dynamism practice, investing in seed through growth stage companies. a16z recently completed a new $15B firm-wide raise, and now the American Dynamism fund totals nearly $1.8B.
The practice has invested in standout companies, including Anduril, Saronic, Hadrian, Northwood Space, Radiant Nuclear, Base Power, Eperus, Flock Safety, CX2, and ASI.
Before Andreessen Horowitz, David founded OpenDNS, a leading internet security company acquired by Cisco. David, welcome to the show!
0:58
Hey, Nick, thanks for having me. Yeah, it’s good to see you again, sir. So tell us you know your two minute backstory. You know, what is your path to becoming a VC? Yep, I am a lifelong entrepreneur and a builder. I built a cybersecurity company called open DNS. It took, you know, as as is the case with many startups, it took 10 years to have an overnight success. We built that company halfway. It was originally a consumer cyber security company, five years in, pivoted to being an enterprise cyber security company, sold that to Cisco. I had never been a I never served our country. I’m not a veteran, but at Cisco, I got my first exposure, both to working with lots of veterans, but also to working with the government as a customer, both across the intelligence community, the Pentagon and the federal government, and really came to respect and appreciate the men and women who work in government or serve Our country, and that that was my first exposure, and that ultimately led, after three and a half years at Cisco, I ran the global security business, I stayed a year and a half longer than I had to. I had a two year commitment there, but I stayed three and a half years. Really enjoyed it, but ultimately wanted to do something a little bit different. Didn’t want to start a company I knew mark and Ben from when they started the firm Andreessen Horowitz, and they actually passed on investing in open DNS, but did it very nicely, and I stayed in touch with them. And so when I was getting ready to leave Cisco, another venture firm actually sort of recruited me and got me interested in investing, while simultaneously convincing me that I would not enjoy their firm, which we could talk about, if we want, about how every firm is different, even though, generally what we do is a commodity. We sell cash for equity. But Ben and Mark encouraged me to spend some time with the firm, and ultimately, Ben said that if I joined the firm, I could do whatever I wanted, which was a very good pitch. And so in 2018 I joined the firm, and a year later, I made my first defense investment, which paved the way to the American dynamism practice and now set of funds so.
3:08
So how does a firm recruit you and explain to you that you would not be a good fit for their fund? How do you square that?
3:16
Well, okay, so another firm, which will remain nameless, you know, asked me to come in to spend some time with them. One of the partners showed up 30 minutes late, and during the entire interview, talked about himself, literally, I don’t think he asked me a single question. Oh, my, that was like, not a great experience. They also, you know, I, even though I never worked with Andreessen Horowitz, you know, Ben was such a prolific writer, and Mark, Mark wrote these long essays. On occasion, I always sort of believed in what they wrote. I certainly believed in Ben’s view that a venture capital firm should do more than just sell cash for equity, and in fact, that the management fees should be especially when you’re at a large, scaled firm, the management fees should be used to build out support services for a company to help out in moments that mattered. And I’d had my own adventure with VCs, with my own company, some some very positive, some less positive. And so I believed in the idea of Andreessen Horowitz and this firm was like the polar opposite. This was, you know, five people going into an office, sometimes working as lone wolves, doing their own deals, no investment committee, very boutique, artisanal view of venture. And I just felt like venture was evolving, and that wasn’t the model for me. So I didn’t, I didn’t really like the culture of that firm, and I didn’t really believe in the philosophy of that firm, but they did get me excited about working with entrepreneurs. That part worked, and then that’s great, you know? And Thoreau had sort of stepped into the picture at the exact same time. And the guy who runs GP recruiting at the firm, I think he has calendar invites whenever, like a founder, sells a company. I think he puts a calendar invite in his calendar to reach out to them the next year, on the anniversary, and the year after that, the year after that, because he had reached out to me every year for. Three years. And finally, on the third year, I was like, All right,
5:02
I’ll talk to you guys, amazing. So you went through this experience where you started open DNS in the consumer space, you switched to enterprise. Now that you’re an investor, how do you handle a really compelling startup, great product, great team, but maybe they’re in consumer and you feel like they should shift to enterprise. Like, how do you do that analysis? And do you, you know, recommend that they make the switch while you’re, you know, in a in a diligence process?
5:35
Yeah, well, I think you know this great team, bad idea or great team, unclear. Go to Market is always better than a bad team. With a great idea or a bad team, you know. So, so when there’s a great team and all the ingredients seem right, but there’s some key ingredient that hasn’t quite been unlocked, where they’re trying go to market. You know, in my case, consumer versus enterprise. Or we actually think sometimes we invest just when the idea is not so good, but the team is just extraordinary and has expertise in the area. We’ll write the check, especially early on, on a seed bet. I think, you know, when we write seed checks, or even series A checks, we’re still, you know, very much betting on the team before the progress as a parent. You know, in my case, it was a little bit easier. I had built this consumer business, and actually, everybody loved it. We just weren’t making any money. It turned out, consumers really did not want to pay for cyber security. They’d much rather buy a cup of Starbucks coffee. And in fact, lots of businesses were using our product, but also not paying for it. And you know, at some point, you know, you know, you know, you have to be a rocket scientist to figure out, well, maybe we can just charge money to these businesses and so that that pivot from consumer to enterprise sort of happened naturally, but it happened simultaneously with me raising traditional venture capital and really shifting the composition of my investors, which worked out in my favor, because my investors were very supportive of this idea of building an enterprise business. I think when evaluating companies right now, we generally on the seed stage, will make the bet, if it’s a great team bad idea, or great team bad go to market. And in American dynamism, people often think like, oh, I can only sell to the government. But oftentimes, it turns out that there are often commercial applications for a lot of technology, or the opposite. Somebody will pitch us a commercial idea, and we think, well, actually the customer is primarily the government. In the case of radiant nuclear, where I’m on the board, when we first invested, we thought this was really a great replacement for diesel generators powered by nuclear power. There is a huge commercial opportunity, but there’s a lot of regulatory challenges that come with that selling to the government is much easier path, and I think now will be the primary, dominant path for the foreseeable future, and then we’ll expand our commercial offering.
7:51
Yeah, what do you think about startups that pitch you, let’s say early stage, seed stage as dual use. You know, we’re going to sell the government and we’re going to sell, you know, commercially, in an enterprise like, should that be a staged process? Should it be phased? How do you figure out which one to do first? Like, walk us through your thought process on that.
8:11
David, I personally prefer, and I think most of my partners agree with this. Prefer when companies are not dual use at the outset, I think founders pitch us dual use because they think that it’s more attractive to us. They think that, you know, maybe the government’s not a big enough tam or too hard of a sales cycle, so we’re going to be turned off. So they need to pitch us this dual use thing. But the thing is, it’s hard enough for a company to succeed in one sales go to market swim lane. They’re like, oh, I want to add two at the very beginning. When I have no sales organization, I have no sales leadership, and I want to prosecute both of these things, but that’s even harder. Yeah, you know, it’s hard enough to not die with one go to market motion, so I generally prefer when companies just pick one. And I’m fine with it being the government. You have to remember, the government is not one buyer. The government has many 1000s of buyers. You know, if you say, even in the Pentagon, if you sell to the Special Operations Command, that’s different than indo, PAYCOM, different than the Air Force, different than Space Force. You know, they’re all different organizations. You sell to the CIA that’s totally different than selling to the NSA, totally different than selling to, you know, Homeland Security or some other organization inside the government. And they’re all very different buyers with different sales cycles and buying processes and procurement processes. So we prefer when they focus on one and try to do it exceptionally well. We, you know, and then as they mature, and they and they get more reps under their belt, and more maturity, or more mature, more maturity organizationally, then it’s fine to add a commercial swim lane, or vice versa.
9:37
So should founders when they’re selling into government and they’ve got all these different buyers of shapes and sizes as as you articulated. Some have different sales cycles. Some have different stakeholder committees and decision makers and users like, how, how does a founder navigate? You know that process in the early stages, when there aren’t really. Really great, established playbooks, you know, selling into special ops versus the CIA, yeah.
10:07
I mean, we’re very lucky. It’s very hard, but we’re very lucky that it’s easier today than it was five, 710, years ago, because companies like Palantir and even to an extent Now Andrew and certainly companies like SpaceX, you know, they all had to sort of crawl so that, you know, other companies could walk, and now new startups can run, I would say, when it comes to government, go to market, and there’s a lot more educational material out there. There’s people like Steve Blank, who’s a professor at Stanford, who teaches a class called Hacking for defense, and he has published prolific materials and selling to the government, and how you find an acquisitions officer and a program office? How do you go through the appropriations process? And you have to remember selling to the government, certainly at the large dollars, you know, not not these small research contracts, but the large dollars you need, the money appropriated by Congress. And then you need to still find a buyer to go actually procure and buy you and select you, and you have to fit within their selection criteria that can often include an RFP process or a bake off process. And so it is hard, but it is certainly knowable, and there’s more people today in the startup ecosystem who have gone through these sales cycles and sales processes that you can bring into your company. You know when Palantir and SpaceX were starting, both companies had to sue their customer, the government. You know, SpaceX very famously sued NASA. Palantir has also sued the government. Nowadays, I don’t think startups necessarily need to sue the government to get, you know, to be commercially considered for a solution. And there’s just a lot more capability. That said it’s hard, but it’s also hard to be a founder selling enterprise software who’s a 22 year old has never worked in Title large enterprise or never done enterprise sales. And you know the same way you read books. You read benioff’s Behind the Cloud, you read software about the story of Oracle, and you learn how you know real enterprise sales forces and teams are built. You read about how SaaS companies were built to do enterprise sales, and then you you decide what makes sense for your business. And the same is true with selling to the government.
12:10
So many years ago, I worked in aerospace, and I distinctly remember hanging out with a leadership team. We went golfing, and one of the top sales guys, you know, he was just a guy. He was just a good guy. We played golf. He was friendly, you know, he was, he was, he was a lot of fun to be around. And after the meeting, I remember looking through our financials, and he was the highest paid person on our payroll by far, like he was making 3x what the CEO made. And I remember asking the CEO, why is Steve getting paid 3x what you get paid? How does that make sense? And his reply to me was, Well, Steve, you know, sold that contract on the new 787, program. To you, it looks like a $2 million contract, but that’s a 20 year program, and that’s going to be, that’s a $40 million ticket, and that’s just what he did last month. So he, by far, is driving, you know, the most revenue for the business of anyone, and that’s why he’s the highest paid. And so my question to you is, like, as you think about these long term programs in government, you know, programs that once you get specked in, especially if it’s, you know, single source. But in some cases, it might be dual sourced. When you get specked in that’s like a completely different calculus on net revenue retention, you know, if you’re the supplier on a program that’s going to last for 1520, years. So how does that factor into kind of the investment process and the analysis process, as you’re validating, you know, future revenue streams.
13:46
I think every you know, I don’t know, day, week, month that we spend working on the American dynamism practice area, and what we’ll talk more about what that is a little bit. But in these let’s just use defense as one, one component of it. We realized that we just have consistently underestimated the size of these contracts. And you’re absolutely right. There’s sustainment programs. A lot of these programs expand over time as they get more penetration. You start in Special Operations Command, you expand into CENTCOM, you expand into all these other different groups within the military, by the way, you then get a domino effect. And in many products, you can then do foreign military sales. And you know, when we invest, we don’t really think about, you know, what the UK Royal Marines might do, or the UAE government when it comes but they like buying our products. They like buying our defense products. And so I think we just consistently have underestimated how large and how quickly these contracts can grow. Now that said one, one downside, which I think we have also come to appreciate, is that while sales is very hard and getting these contracts is hard, you also have to get into production. You have to build the factory to make the thing. It’s not just like a software company or enterprise SaaS company that can commit the code and ship the software and. And once you build, commit the code and have a release, you can go sell, you know, an infinite number of copies. If you are making missiles, you have to then go actually make the missile. And so it’s one thing to get the contractor that is hard, and we recognize that that’s an achievement, but then you have to deliver the product before you can recognize the revenue. And so we do look at our companies and evaluate them, not just on the contracts that they’re able to manifest or bring on board, but then what is the lag before you can actually start to recognize that revenue and deliver a successful product to the customer? And we see companies like Andrew are very fastidious. They’re a multi product company. They’re very fastidious about measuring contracts to revenue delivered, and the production lag between those things across all their SKUs, because at the top line you can you can obscure that. You might have one product where you know you get a contract that you can deliver immediately. You might have another product where you can’t deliver for 18 months because you have to set up a whole supply chain and manufacturing and factory line to get the product out the door. And so these companies look really good on the top line, but when you come to the recognized revenue can be very different. And so our best companies measure that methodically and are relentless about really staffing up their production capabilities. Because, you know, getting the contract is great. Steve gets paid the sales guy tip of the spear, but, you know, the company doesn’t get paid unless they can recognize that revenue and and we see that that’s a dimension that we just haven’t seen before in other venture backed companies that that I think we have more institutional knowledge of working with. I mean, it’s almost
16:34
like the founders have to set up a separate company, right? It’s like they have their R and D tech company, but then when they get into manufacturing, you know, it’s, it’s a whole different workforce. It’s a facility, it’s inventory tracking. It’s like, that’s when you’re
16:51
dealing, you’re dealing, oftentimes, blue collar workers, which most tech companies don’t have a lot of experience having to deal with. You have to deal with very complicated supply chain dependencies. You know, people talk about agile software development in the enterprise world, there’s no real agile. And a lot of the defensive production kind of work, it’s all very waterfall. If you don’t have the motor that you need for your drone, you cannot make the drone. You can’t just do something else in the other order you need, you need the bushings, and you need the grease, and you need the motor. And you know, if you’re doing something else, you might need the actuator. And where does that part come from, and is that part compliant with the rules you’re selling into? And, you know, Elon famously said, like, the factory is the product, and we just see that our best companies that is, that is a mentality that is totally true.
17:33
So David, we kind of jumped right in. But can you help frame, like the American dynamism, dynamism practice what it is you do and how you support, you know, founders, because there’s a lot of talented software engineers that are embarking on these, you know, heavy production component businesses. And so talk to us about that aspect as well. Yeah.
17:54
So the American dynamism practice, and now funds is, is an investment thesis focused on investing in areas that serve the national interest, which is a broad term, but broadly means anything that makes the American dream more real, which for us is people want to own a home in a good neighborhood with good schools, a short commute, in a safe environment, in a country that is safe and able to defend itself. And so for us, those categories are often defense, energy, public safety, space. It can include transportation, logistics, it can include the supply chain and manufacturing. It can include things. We would include things like healthcare, except that we have a whole team that does healthcare, so we stay away from it. Obviously it’s super important to the national interest. But I would say defense, energy, public safety and space are the primary categories. And then, you know, oftentimes the companies we invest in sit at the intersection of hardware and software, usually commodity hardware with very advanced software. And you know, we’re in a moment in time where we have rapid advancements in technology from Ai, autonomy, computer vision, robotics, automation, that are transforming a lot of these legacy a lot of these categories and the legacy players are really being pushed by the side as we transform that both what the customer says through our needs on this, on the demand side, but also the way that we supply them. With advanced manufacturing, with new technologies, commodity hardware, advanced software makes us, from our standpoint, a super interesting time to invest in these areas, a much higher margin set up, set of outputs in these categories that traditionally have been very low margin. So we’ve been doing this for now more than five years with a name, and a little bit longer than that. I mean, the firm has been doing this for a long time with no name, and then you asked about the teams that we build to support these companies.
19:44
So happy to jump into that awesome, awesome, love it. I guess you know, another thing that I gain an appreciation for from the aerospace days is that you have component manufacturers, you have subsystem manufacturers. You. System manufacturers. Then you have tier twos, tier ones, primes, above that you’ve got, you know, the prime might be Hamilton, Sun strand, the OEM might be Boeing, right? And then finally, you’ve got the carrier, which is United Airlines, right? So there’s all these layers to the stack, and I know that mark and Ben have wrote about fat startups and vertical integration a lot lately, you’ve talked about shifting left. So thinking about all the components of the supply chain, you know, motors, PCBs, subsystems, talk to talk to us about how this factors in to the investment decision. You know, there’s a lot of risk throughout that supply chain. So are you looking for vertically integrated startups that are trying to build it all themselves, like, you know, Elon’s methodology, how do you assess fragility in that stream? You know, because if there’s only one component manufacturer for a particular specialty motor, you know, that presents a lot of risk for an end product. So how do you think about this?
21:01
David, yeah. I mean, we generally prefer to invest in companies that are much more full stack. So you know, if a company pitches us and says we’re making a very small optical laser that has this use case, and it’s very narrow use case, but over time, we’re going to expand that could be interesting, but it may not be so interesting for us. It doesn’t mean that they’re not even interesting in the venture landscape. They’re just not interesting at our scale. You know, our latest fund is almost $1.2 billion our previous fund was, it was about $600 million and so for our scale, we want to make big bets in companies where we think that the TAM really is sort of unbounded, or at least unbounded in the in the near to medium term and and we want companies that can control their own destiny and not be so reliant on supply chain requirements. So you’ll look at companies like Android, where they will hire, they’ll they’ll buy their own CNC machines. They’ll buy their own, you know, Metal Bending machines. They will assemble their own components where they can and on components where they don’t yet have capability, they’ll try to build a diverse supply chain. And there are, there have been examples where we’ve learned, hey, wait a minute, a bunch of our companies all need brushless motors that requires magnets. Are there? Is there a company out there that is going to build motors and actuators that can actually supply all of our companies and as the US increasingly makes determinations that certain things need to be made in America or made by our nearshore allies, things like motors and actuators become an investment opportunity. So we actually have, I use the example of like the optical laser. Maybe that’s not a perfect fit for us, but then there’s others examples where we think that these component suppliers actually can serve a whole set of constituents that we that we’ve invested in. And so we’ve invested in a company called West mag. They make motors and actuators. I think that’s public. We’ve invested in them, I hope. And but they make them here in America, and they will be a supplier to many of our companies, and over time, they will add more and more components that they manufacture. We are investors in a company called the Advanced Manufacturing Corporation of America, Amca. AMCA is actually going through that subprime and even sub sub prime list of suppliers and component makers, many of whom are, in fact, in the US. They’re aging out. They’re retiring. They don’t want to be in business forever. They’re very, very good at a low mix, high volume set of components. And am because coming in and either buying these companies and introducing new software to optimize their processes or transform the way they manufacture and is going to try to build the next generation of like a transdyne or this Mom and Pop component supplier. So we think there is lots of opportunity in the like sub subs, but for the big bets we make, they tend to be these full stack companies like an Android, like a Saronic.
23:45
So David, I read the Andreessen blog post why America needs dynamic defense reform. In it, you’ve outlined five, five major reforms, from commercial first, procurement to workforce incentive changes, but these have to move through the NDAA process, which, as you know, can take years for your fund.
24:09
Now, yeah, we’ve now made progress. Four out of the five, I think, have been put into the NDAA
24:13
No kidding, all right,
24:14
yes, very good. Forget which one didn’t get included, but commercial first certainly is included,
24:21
unbelievable. So the policy work is paying off,
24:25
yeah, well, I guess we asked. We have a whole team in DC that, you know, not not to, like, make a sales pitch, but, like, we have a whole team that works on government affairs, and we have people that that work through the House and the Senate to really inform them about what startups need to advocate on behalf of startups, which I think is a rising tide for all companies, you know, but we, we generally try to give our companies an unfair advantage. But a lot, a lot of that work is being pushed through, and there’s an administration that, you know, frankly, is very friendly toward this idea of transforming the procurement process for for the Department of War.
24:54
Okay, very good. Well, congrats on the four or five. Hopefully you can get the other one across the line. So. Um, you’ve also emphasized that production is deterrence. And talk about reigniting American manufacturing with new factories and production capacity, but building factories and scaling manufacturing requires enormous amounts of capital and longer payback periods, which does not sound like maybe the most recent venture model, you know, plowing money into SaaS. How do you square that circle? Right? Are you actually funding the manufacturing side? Are you focusing on software and enabling technologies that sit adjacent to production?
25:35
Yes, and yes, it depends on the company. Depends on the stage as companies scale up. Now, one thing we’ve seen is that we’ve a lot of our companies now can unlock downstream capital. So major equipment financing, and we’re not talking about buying, like, you know, 20 laptops. We’re talking about equipment that costs three, four or $5 million a pop, has huge resale value. Those things now become much, much more easily financed. We’re seeing project financing for large programs and projects where you can essentially get huge advance lines when the government commits to spending $400 million with you, you can go get $200 million to go build your factory from from other financial partners. You don’t have to use venture dollars for that. But early on, we are financing it, and we think that we’re doing it in a way and with companies that will have a higher margin profile, faster time to production than what has historically been seen in the past. In fact, many of our companies are picking geographies, you know, where they’re not in California or places that have way too much regulatory red tape. They’re moving their factories to Arizona or New Mexico or Tennessee, where the government says, We will expedite your permits. We will make sure that if you apply for a permit, you will get an answer in 24 hours, and if you need help to resolve it, we will be your partner. We will work with you on your permits. We will give you power. We will provide you the power that you need to run your factory, and we will help you hire the people who are often you know, either unskilled or highly skilled, but blue collar workers. And so, you know, I’m channeling Chris power from Hadrian here one of our companies where, you know, we build, we encourage our companies to build factories in places where permits, power and people are all readily available and possible. And so we do invest in these companies at that stage, but there is a lot more downstream capital that’s now available as they scale up and prove that they can deliver but we certainly have to bootstrap it, and we’re happy to do that. You know,
27:29
we’ve talked a lot on this program, and there’s been a lot of headlines lately about AI sovereignty. Something that feels relevant is a certain degree of industrial sovereignty. You know, for a nation, you know, what do you think is missing today, sort of in the US innovation stack, when it comes to, you know, critical technology that’s important to our future, but we may not have, you know, the capability within our borders to to innovate it and produce it.
28:02
Well, if I went through the whole list, it would be very it would be very depressing. And I’m an optimist, and I think you’re an optimist. So we, we won’t go through the whole list of all the things that we are totally dependent on China or people that are across large bodies of water to supply to us. What I will say is that when you think about a missile that has a guidance system, has a gyroscope, has rocket fuel, generally solid rocket motor, although increasingly liquid, liquid fueled rocket motors an explosive capability, you know, some kind of a metal housing around it, there are a lot of those components, from the chips to the fiber gyros and things that come from overseas. And if we were in a conflict and we decided that we needed to make 10,000 missiles, we can be sure that even if it was from an ally, even, let’s say it comes from South Korea, the gyro if we’re in a major conflict with China or in the Indo, Pakistan region, I don’t think any country like South Korea that makes gyros is going to start sending them back to us. They’re going to keep them for themselves. So we need, as a matter of national security, the ability to manufacture a long list of components, batteries is a key one. We are entirely, not entirely. We are largely dependent on China for battery and in fact, we’ve seen some of our companies get sanctioned by China and are no longer able to buy batteries from China, and it’s a very good preview into how painful it is when you’re prohibited from buying batteries or motors or actuators from China. And so having near shore or domestic either way is fine. Can be Canada, can be Mexico, can be even South America could even potentially be Europe, having really strong allies who we actually have in a supply chain that we can reach in a conflict, be able to supply us these key things from rare earth materials and magnets, motors, actuators, chips, PCB assemblies. We don’t need to make every semiconductor chip. It’s okay that Taiwan still. Makes a lot of the very, very small chips that we rely on for AI compute, but there’s lots of chips that go into missiles and guidance systems and other programs that we could be making here, and we increasingly are. And then, if you look at even the pharmaceutical supply chain, all the chemicals and precursors we need, even things like saline. We found out during covid, a lot of our saline comes from China. A lot of it comes from Puerto Rico. And if there’s a hurricane in Puerto Rico, and we’re at odds with China, we will have no sailing in this country, which turns out that sailing is really important for anything you do in a hospital constantly. And so we need there is a descending list of priorities, and I think the government is now quite vocal about what some of those priorities are that we need to remediate. Some of them are opportunities for venture capital. Some of these are just things where the government needs to step in. I don’t think there’s a great business to be built in Saline production, but the government absolutely needs to do something to motivate saline production for brushless motors and actuators and things like that. There might be a great venture backed opportunity there for lithium extraction, for battery manufacturing, for battery recycling. There might be great venture backed opportunities there. So I’m very optimistic, but there’s a long list of supply chain dependencies that we have and that we need to resolve in relatively short order. So David,
31:16
you’ve you’ve described space as a core theater for deterrence. Of course, we have a new military branch, the Space Force. Now, where do you think the US stacks up against other major powers with regards to space dominance?
31:33
Well, I think space thanks to SpaceX, we went from being at risk of losing our dominant position to being unquestionably in first place. I think it’s important that we talk about not just launch capabilities, but full space domain awareness, which means, do we understand what’s happening in space and what other people are doing in space? Do we have space domain dominance? Meaning, can we put things in space to either communicate or observe what’s happening here on earth with reliability, regularity and high fidelity. SpaceX has really solved the launch equation for the United States, for both the government and for the private sector. I believe that SpaceX has launched more government satellites in 2025 than all government satellites launched cumulatively in history before that. So they certainly have solved the launch problem there. You know, we increasingly see that things that we thought were disparate, like, you know, defense, energy and space are actually very interrelated without very strong space domain dominance and awareness, you don’t have reliable national security and defense. You need it for communications. You need it for observability. You need it for ISR. You it is an important area. The other thing SpaceX, and where we are today, is that SpaceX is unlocked, is that you can now put things in space. If you’re a private company, if you have a new optical laser communications device or a radio communication device, or an imager, or some kind of Earth observation capability, you can now get that in space relatively quickly, and that unlocks a whole new economy. And I think the next thing to unlock will be when we can bring things back from space reliably, which is not currently possible. It’s easy to put things in space, it’s very hard to bring them back, but once, once we’re able to bring things back from space, you know, to me, it feels kind of like the new world where once we were able to travel very reliably between Europe and the New World, it unlocked a whole amount of economic expansion in the new world. And, you know, ultimately give rise to the United States. And I think space could be equally as exciting once we can manufacture in space. And you know, that’ll be the next big unlock in space. Launch has covered. Return to Earth, I think, is the next big one, and along the way, we cover everything from space domain dominance from a national security standpoint and space domain awareness. I mean,
33:49
it feels like such a critical thing to be dominant in the future. It feels like the next frontier that that needs to be controlled. I mean, AI could be in that category. There’s, there’s some other things, maybe quantum but Elon
34:05
now talks about space data centers, which, I think, yeah, you know, at first sound crazy, sounded crazy. Now smart, right? It does sound crazy. It still kind of sounds crazy. But now smart people are like, well, you know, there’s certain scenarios where the math works out and, you know, maybe it’s possible, and there’s, there’s other people we talked to that think there may already be some space data centers already up there, and, you know, so we may have more exposure and experience of how this is already working out than people realize. Like, Elon’s a smart guy. He talks to smart people, and then on the communication side. I mean, we know that things like Starlink are transformative. And if you look at a country like Taiwan that has really two primary entry points for subsea fiber cables. If those get cut, they would be offline, if not for starlight, which can provide continuous connectivity for the
34:49
for the country, who’s number two or number three? And this, this the new space race. I guess.
34:55
I think the US is number one. I think China is number two. And I’m not sure that there is. Number three, right now, okay,
35:01
so Russia or India or not,
35:04
I think Russia has really fallen off. I would have to try to check on that. I think India is still fairly nascent. And when they do a launch, it’s largely, I think, powered by us technology. I think France does still launch, but in India might be there, but I think nobody, I think it’s like the power law, like in venture. I think number one is, you know, the US, by a wide margin, China is number two. And then I think it’s a very, very
35:26
steep drop to number three. So we’ve been talking about space, you know, as you look to kind of the long term future for the American dynamism thesis and practice, you know, what are, what are the other industries in your umbrella that you’re really bullish on?
35:39
Yeah, I mean, I think that shift left idea of what are all the things that are needed to have a reliable manufacturing base and defense industrial base? Are there housing construction technologies or materials that could be interesting? Is it around, you know, if we’re going to have this robotics revolution, are we going to be building parts of those robots here in the US? Do we need more about, you know, what do we need to make batteries? We need lithium. We need copper. Are there mining companies that could take us off our first approach, which, you know, again, maybe sounds crazy on the surface, but maybe actually makes a lot of sense, that then the sort of like the blindly crushing rock methods that historically had happened in mining. And so I think that shift left is one, one aspect. The other is when we think about energy, you know, we are in a country and in a world that has an insatiable thirst for energy, whether it’s aI data centers or electric vehicles or just the general electrification of our of our lives. So anything that has to do with power generation power distribution to make our grid more reliable and more stable or and then power storage are all becoming incredibly important. And then I think, you know, I don’t focus too much on cyber security, but it’s very clear, if you look at what happened in Venezuela with Maduro, that cyber was a major part of that campaign, even from what’s just been publicly reported, that’s clear. And so having very, very clear capabilities, both offensive and defensive in the cyber domain, I think are incredibly increasingly important to national security, and certainly fall under our remit, at least also because of my background 100%
37:18
just to double click on your energy point. How do you think about, you know, there’s investments going into nuclear, significant investments in uranium. You know, solar has now dropped to, you know, some of the most economical levels we’ve seen. Natural Gas, of course, is prevalent here in the states, like, how, how do you figure out which energy categories are productive uses of capital? You know, when some are heavy capex, they’re really long term, like nuclear, and then we have some that, you know, the economics have gotten quite good, like solar and can be deployed fast, and natural gas can be extracted. So how do you think about investing in the energy segment?
38:08
I like that you didn’t say wind, because wind is probably the worst, but I agree solar is very attractive, and there’s lots of times now where, you know, it’s just clear that deploying solar, especially combined with batteries, is now reached a cost part of the, you know, a point on the cost curve where it makes absolute sense. I’m not sure that there will be major solar advancement from a efficacy standpoint, where it makes sense for a venture investment, although I think people in the venture community have made money in solar investments, but we we have relatively limited exposure there. And I think on nuclear, it’s the same thing. We look at the cost curve of can we do nuclear at a price point that makes sense? What needs to be true for that to happen? Does the regulatory environment need to change to make the regulatory cost of building a nuclear reactor feasible? Does the cost of diesel need to go up? You know? So when we made the investment in radiant, radiant radiant nuclear, for instance, you know, we sort of evaluated that when we thought diesel might cost $8 a gallon. But on the flip side, we look at how much the US military spends on moving, transporting fuel to forward operating bases, they can sometimes pay $200 a gallon. So at $200 a gallon for diesel fuel, all kinds of power start to feel much more attractive. In fact, almost every power system and a generation system can come up with is much more attractive at $200 a gallon. And so then nuclear power starts to become make much more sense. Then you have to look at, well, do you have the team that can do it? Do you have a regulatory environment? Do you have a fuel supply chain that can deliver and so for radiant that all felt like it made sense for us. For us. You know, we’ve, we’ve avoided making nuclear fusion bets because it feels too early on, the sort of the technical readiness level, the TRL scale, where we try to avoid, we think building a company is hard enough that we try to avoid the science projects. You. And so we haven’t made those bets. But, you know, it comes to nuclear fission, we know how it works. We know it’s possible. And I think that it’s now very bipartisan tailwinds in the regulatory environment and the government standpoint to approve and generate more nuclear power. And then, you know, from it, from a just like a broader standpoint, outside of investing, I think we have to do all those things, natural gas, nuclear, solar, we need all the more batteries. It’s not like we can just do one and not the others. Our thirst for energy is so insatiable that we could do all of them, and it would not be enough.
40:29
Is it realistic that Elon’s plan to power the data centers in space with solar, which will be much more effective because it’s closer to the sun and doesn’t have to deal with atmosphere, etc. Does that feel like it’s past TRL, like it, like it’s viable.
40:43
I think, I think the question is, like, how do you radiate the heat in space generated by the data centers? But on solar, you know, we can deploy a one kilometer wide solar array in space. We don’t have to worry about land or dust or somebody stepping on it. You know, they are very capable of deploying very large solar arrays that are more efficient in space and super reliable. I think that makes total sense, and it’s very, very feasible. My understanding, and this is kind of the extent of my knowledge. Here is the real issue is, how do you radiate the heat? Space is both incredibly cold and can be incredibly hot, and when you generate heat because there’s no convection, radiating that heat away is quite challenging, interesting from from what i Okay, David,
41:27
a few quick ones to wrap up here. What book, article or video would you recommend to listeners?
41:34
Oh, I should have come prepared with this one. Well, I did. This is sort of lame, I guess, but I just did read an essay somebody wrote on x about how the recent release of Claude 4.6 and the recent release of chatgpt when it comes to code generation, really is a watershed moment. You know, I feel like every six months we’ve been in this time where it feels like, wow, AI is incredible. This is really changing everything, but maybe now really feels like that time, and it just makes me think, well, 18 months now, it’ll be even crazier, but, but the recent releases from anthropic and from companies like cursor that use some of the other llms really do feel transformative. And that like, like, when I was a mediocre to moderately good engineer, literally everything that I could have ever done now can be done by the AI. The AI is, for sure, a better programmer than I am at this point, and I’ve been playing around with it, and it’s I was on DND, except for when my parents called Awesome.
42:40
David, do you have any habits or behaviors that are a secret weapon?
42:45
No, I should exercise more. I have no excuse that would be a better. And I think as people age, my observation, which is not profound, is that the people that exercise and are healthy feel the best as they age, and as I watch people that go into their 60s, 70s and 80s, the people that were probably in the best shape and had good course muscle strength in their 40s and 50s, I think, are doing the best. So I
43:10
intend to be more like that. Love it. And then finally, here, what is the best way for listeners to connect with you and follow along with American dynamism. Yep.
43:19
D you at a 16 z.com for my email, and I’m on x at, David, you at, D, A, V, i, d, U, and easy to find. All right.
43:28
He is David ulovich. The firm is Andreessen Horowitz, American dynamism, David, thank you so much for your time today. This was a lot of
43:37
fun, awesome. Thank you, Nick. You Brenna.
43:43
All right, that’ll wrap up today’s interview. If you enjoyed the episode or a previous one, let the guests know about it. Share your thoughts on social or shoot them an email. Let them know what particularly resonated with you. I can’t tell you how much I appreciate that some of the smartest folks in venture are willing to take the time and share their insights with us. If you feel the same, a compliment goes a long way. Okay, that’s a wrap for today. Until next time, remember to over, prepare, choose carefully and invest confidently. Thanks so much for listening.
